https://openaccess.leidenuniv.nl
License: Article 25fa pilot End User Agreement
This publication is distributed under the terms of Article 25fa of the Dutch Copyright Act (Auteurswet)
with explicit consent by the author. Dutch law entitles the maker of a short scientific work funded either
wholly or partially by Dutch public funds to make that work publicly available for no consideration
following a reasonable period of time after the work was first published, provided that clear reference is
made to the source of the first publication of the work.
This publication is distributed under The Association of Universities in the Netherlands (VSNU) ‘Article
25fa implementation’ pilot project. In this pilot research outputs of researchers employed by Dutch
Universities that comply with the legal requirements of Article 25fa of the Dutch Copyright Act are
distributed online and free of cost or other barriers in institutional repositories. Research outputs are
distributed six months after their first online publication in the original published version and with proper
attribution to the source of the original publication.
You are permitted to download and use the publication for personal purposes. All rights remain with the
author(s) and/or copyrights owner(s) of this work. Any use of the publication other than authorised under
this licence or copyright law is prohibited.
If you believe that digital publication of certain material infringes any of your rights or (privacy) interests,
please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make
the material inaccessible and/or remove it from the website. Please contact the Library through email:
OpenAccess@library.leidenuniv.nl
Article details
Groeneveld I.F., Pas S.L. van der, Meesters J.J.L., Schuurman J.M., Meijeren-Pont W. van,
Jagersma E., Goossens P.H., Kaptein A.A. & Vliet Vlieland T.P.M. (2019), Illness
perceptions of stroke survivors: Predictors and changes over time – A 1 year follow-up
study, Journal of Psychosomatic Research 116: 54-61.
Contents lists available atScienceDirect
Journal of Psychosomatic Research
journal homepage:www.elsevier.com/locate/jpsychores
Illness perceptions of stroke survivors: Predictors and changes over time – A
1 year follow-up study
Groeneveld I.F.
a,b,c,⁎, van der Pas S.L.
d,e, Meesters J.J.L.
b,c, Schuurman J.M.
b,
van Meijeren-Pont W.
a,b,c, Jagersma E.
b, Goossens P.H.
a,b,c, Kaptein A.A.
f,
Vliet Vlieland T.P.M.
a,b,c, on behalf of the SCORE-study group
aRijnlands Rehabilitation Centre, Leiden, the NetherlandsbSophia Rehabilitation, The Hague, the Netherlands
cDepartment of Orthopaedics, Rehabilitation, and Physical Therapy, Leiden University Medical Center, Leiden, the Netherlands dMedical Statistics, Department of Biomedical Data Sciences, Leiden University Medical Center, Leiden, the Netherlands eMathematical Institute, Leiden University, the Netherlands
fDepartment of Medical Psychology, Leiden University Medical Center, Leiden, the Netherlands
A R T I C L E I N F O Keywords: “Depressive symptoms” “Illness perceptions” “Health outcomes” “Longitudinal study” “Stroke” “Rehabilitation” A B S T R A C T
Objective: To describe the illness perceptions (IP) of stroke patients in the first year post stroke; to identify
patient clusters with comparable IP trajectories and determine their associations with health.
Methods: This prospective study included consecutive stroke patients after medical rehabilitation. Three and
12 months post stroke they completed the Brief Illness Perception Questionnaire (B-IPQ) and questionnaires on physical and mental health. All eight IP and their changes over time were described. Clusters of patients with comparable IP trajectories were constructed by k-means clustering, with subsequent comparison of patient characteristics. Multivariable logistic regression analyses were conducted to determine the association between IP clusters and 12-month mental health.
Results: Hundred-and-eighty-four patients were included (men n = 107 [58.2%]; mean age 61.1 [SD 12.7]
years). At 3 months, the scores of the IP coherence (mean 3.0, SD 2.3) and treatment control (mean 3.2, SD 2.5) were lowest (best), and consequences (mean 6.1, SD 2.8) and anticipated timeline (mean 6.0, SD 2.7) were highest (worst). At 12 months, the timeline and treatment control scores had significantly worsened. Three clusters of the trajectories of IP were identified, and designated as ‘favourable’, ‘average’, and ‘unfavourable’. The unfavourable cluster was significantly associated with worse physical and mental health at 3 months (unadjusted) and de-pressive symptoms at 12 months.
Conclusion: Stroke patients' IP partly changed between 3 and 12 months post stroke. Patients with an
un-favourable IP trajectory had a higher chance of depressive symptoms at 12 months. Illness perceptions could be considered as an additional target of treatment.
1. Introduction
Worldwide, stroke is one of the leading causes of disability and mortality [1]. Stroke may lead to impairments in functioning, limita-tions in activities, and restriclimita-tions in participation, and may affect health-related quality of life (HRQOL) [2]. Within the first years post stroke, a third (pooled prevalence 31%) [3] of stroke patients develops depressive symptoms and a quarter (pooled prevalence 23%) reports to suffer from anxiety [4]. Health outcomes after stroke are influenced by sociodemographic, clinical, treatment-related, and psychological fac-tors [5,6].
As to the psychological factors, post stroke outcomes can be influ-enced by one's perceptions of the illness [7–10]. Stroke patients with negative (‘maladaptive’) illness perceptions (IP) had more distress [7], lower medication adherence [8,9], and more difficulties returning to work [10]. The concept of illness perceptions (IP) is a major component of the Common Sense Model (CSM), stating that a patient forms mental representations of symptoms and disease, in order to make sense of and manage the illness [11,12]. Illness perceptions comprise the patients' beliefs about the symptoms, duration, cause, and consequences of the disease; the perceived emotional impact; his/her concern and under-standing, and his/her beliefs about the controllability of the disease
https://doi.org/10.1016/j.jpsychores.2018.10.019
Received 3 July 2018; Received in revised form 11 October 2018; Accepted 31 October 2018
⁎Corresponding author at: Rijnlands Rehabilitation Center, Wassenaarseweg 501, 2333 AL Leiden, the Netherlands.
E-mail address:igroeneveld@zinl.nl(I.F. Groeneveld).
0022-3999/ © 2018 Elsevier Inc. All rights reserved.
[12,13]. Patients do not hold IP in isolation, they are part of a schema [14]. Patients change and update this schema based on incoming in-formation and subconscious hypothesis testing [11]. Illness perceptions appear a modifiable target of treatment; personalised counselling in-terventions were shown to be effective for patients with myocardial infarct and diabetes [15,16]. In order to measure IP in the CSM context, the Illness Perception Questionnaire (IPQ; 1996) was developed [17]. In 2002 a revised version (IPQ-R) was published, containing additional scales13. As there was a (clinical) need for more rapid assessment of IP, a brief version of the IPQ-R was developed. The Brief IPQ (B-IPQ) contains eight items, all having 0–10 scales [18], concerning perceived consequences, timeline (acute-chronic), personal control, treatment control, identity (symptoms), concern about the illness, coherence of the illness, and emotional response. The ninth item asks the patient to list the three most likely causes for his/her illness.
In patients with stroke, research on IP is scarce. In a descriptive study one year after stroke (n = 15, mean age 73.3, 73% female) on average the participants believed they understood their disease rela-tively well (coherence), experienced relarela-tively low emotional impact, but experienced moderate personal control and believed that treatment was only moderately helpful (treatment control) [19]. In a longitudinal study on stroke patients' and caregivers' distress, patients (n = 42, mean age 65.1, 43% female) experienced relatively few consequences but reported little coherence and low personal control. Also, they foresaw only mod-erate effects of treatment, this IP becoming more negative between 10 days and 3 months post stroke [7]. No studies have yet described the patients' perceptions of the causes of stroke. The course of IP over time as well as the associations with patient characteristics and health out-comes have not been explored either.
In other patient groups the concept of IP has been more extensively investigated, both descriptive and in relation to patient characteristics and health outcomes. Predictors of maladaptive IP included female sex [20], disability, multi-morbidity [21], fatigue [22], depressive symp-toms [20], and anxiety [23]. Changes in IP over time were demon-strated in patients with obesity, chronic obstructive pulmonary disease, heart disease, and osteoarthritis [24–26], and appeared unfavourably related to disease progression. In the cluster of oesophageal cancer patients in which IP became more negative over time, levels of anxiety increased [26]. Illness perceptions have been object of intervention, aimed at adjusting unhelpful IP into more adaptive ones. In patients with cardiovascular disease, in-hospital counselling beneficially influ-enced coherence, concern, and beliefs about the causes of disease [15], and resulted in fewer symptoms and a higher rate of work resumption [27]. Patients with type 2 diabetes reported better personal control, a better understanding, an increased belief in treatment effectiveness, and experienced fewer symptoms, lower levels of concern and distress, and better medication adherence [28]. The mediating effect of IP in the relation between intervention and depression was shown by Jonsbu et al. (2013) in a 3-session CBT programme for patients with non-car-diac chest pain and benign palpitations: A change in illness concern mediated about 40% and a change in personal control mediated about 50% of the change in depression [29].
In summary, IP are related to a range of health outcomes and have been influenced by personalised interventions in other disease cate-gories. Exploring the course of IP and the relation with physical and mental health in stroke patients will provide insight into the opportu-nity and necessity of targeting IP as part of rehabilitation treatment, additional to medication and cognitive behavioural therapy [30]. We conducted a prospective study in a large sample of stroke patients who underwent inpatient specialised medical rehabilitation, and aimed to: 1) describe the IP of stroke patients and their course between 3 and 12 months after the start of rehabilitation; 2) identify clusters of pa-tients with comparable IP trajectories and determine their character-istics; 3) identify the associations between IP clusters and mental health 12 months after the start of rehabilitation.
2. Method
2.1. Design and setting
This study is part of the Stroke Cohort Outcome REhabilitation (SCORE-) project; a multicentre prospective observational cohort study (Dutch trial register no. 4293) currently ongoing in two Dutch re-habilitation facilities [31]. These rehabilitation facilities offer multi-disciplinary rehabilitation treatment including physical, occupational, and speech-language therapy, social work, and consultations with a clinical psychologist [32]. In patients with cognitive, psychological, or behavioural impairments, the rehabilitation team discusses a patients' awareness and emotional impact of the illness, but not his/her per-ceptions thereof. Follow-up data were used from patients who were included in the SCORE-study by March 2016. The study protocol was approved by the Medical Ethics Board of the Leiden University Medical Center (LUMC). All study procedures were executed in accordance with the Helsinki Declaration [33]. All participants signed informed consent. The study was reported according to the STrengthening the Reporting of OBservational studies in Epidemiology (STROBE) guidelines. 2.2. Study sample
The study sample consisted of consecutive stroke patients who were admitted for inpatient multidisciplinary specialised rehabilitation. As compared to the total stroke population, these are relatively young patients, with multiple and complex impairments, and able to return home after rehabilitation. Their average length of stay in the re-habilitation clinic is 1.5 months. For the SCORE-study, patients were invited who were ≥ 18 years old and had an ischemic or haemorrhagic stroke < 6 months ago. Patients with pre-existent psychiatric disorder or dementia, or unable to complete questionnaires in Dutch due to se-vere aphasia or non-Dutch ethnicity, were excluded. Within the first week after admission, patients received an information letter from the treating rehabilitation specialist. Subsequently, a research assistant visited the patient for further explanation. All patients who provided informed consent and completed the questionnaire at baseline were included.
2.3. Data collection
At the start of rehabilitation (baseline), sociodemographic and clinical characteristics were collected using the patients' medical files and a baseline questionnaire. Three and 12 months after the start of rehabilitation, IP and health outcomes were assessed by means of a questionnaire. Based on the participants' preference, the questionnaires were sent by post or by email.
2.4. Outcome measures
Illness perceptions were measured using the B-IPQ [18] For the items on the perceived symptoms (identity), anticipated duration (timeline), and perceived consequences, concern, and emotional impact regarding the disease, a higher score (range 0–10) implies more nega-tive, or maladapnega-tive, IP. For the items on perceived personal control, treatment control, and coherence, a higher score implies more positive, or adaptive, IP. Before data analyses, the scoring of the latter 3 items was reversed in order to facilitate interpretation. The cause item is an open question in which the respondent can indicate the 3 most important perceived causes of stroke, in order of importance. The B-IPQ showed good test-retest reliability, and concurrent, predictive, and discriminant validity in various populations [34]. It was cross-culturally adapted into Dutch [35,36]. It should be noted that for stroke survivors, higher scores on identity, timeline, and consequences cannot be considered ma-ladaptive per se, but higher scores indicate a more negative view of the disease.
I.F. Groeneveld et al. Journal of Psychosomatic Research 116 (2019) 54–61
Disease-specific quality of life was assessed using the 39-item Stroke and Aphasia Quality of Life Scale (SAQOL-39NLg), that was validated for use in people with and without aphasia. The SAQOL-39NLg contains the domains physical (16 items), communicative (7 items), and psy-chosocial (16 items) functioning [37]. Questions are phrased as ‘In the last week, how difficult was it for you to…’ Each item can be answered on a 1 (‘could not do it at all’) to 5 (‘no difficulties’) scale. The total score of each scale equals the mean of the item scores. The cross-culturally adapted Dutch version showed good internal consistency and test-retest reliability, and moderate convergent validity [38]. For this study, only the physical and communication scales were used, as psychosocial functioning was measured with the more comprehensive Hospital An-xiety and Depression Scale (HADS). Considering the phrasing of the items, the SAQOL-39NLg was used as a proxy for physical and com-municative functioning.
Anxiety and depression were determined by means of the Hospital Anxiety and Depression Scale (HADS) comprising two subscales: Anxiety and Depression. Each subscale consists of 7 items that can be scored on a 4-point scale, adding up to a maximum sum score of 21. Scores were dichotomised using the recommended cut-off score of ≥8 to indicate symptoms of depression or anxiety [39]. The HADS scales have good internal consistency, good to excellent sensitivity and spe-cificity, and good to very good concurrent validity [40].
2.5. Sociodemographic characteristics
Sex and date of birth were derived from the patients' medical files. Level of education and living situation were assessed by means of a questionnaire. The patients' level of education was measured using a 6-point scale and split into 3 categories: Low, intermediate, and high education. Living situation was defined as ‘with others’ versus alone. 2.6. Clinical characteristics
Stroke type (ischaemic, haemorrhagic) and localisation (right, left, posterior, stem, both sites) were derived from the patients' medical files. Comorbidities were measured using the ‘Dutch Life Situation Cohort Questionnaire’, a 16-item questionnaire on the most common chronic diseases in the Dutch population, including e.g. diabetes mel-litus, heart disease, rheumatic disorders [41]. The respondent can in-dicate the presence (yes/no) of each disease. The level of independence in activities of daily living upon admission was assessed using the Barthel Index (BI) [42], which comprises 10 items on e.g. bathing, feeding, and mobility; its total score ranges between 0 (worst) to 20 (best). It has an excellent internal consistency (Cronbach's α 0.84 to 0.96) and good inter-observer reliability (Spearman's r 0.77) in patients with stroke. Whether patients still had treatment in the rehabilitation facility upon completion of the 3-months questionnaire was derived from self-report.
2.7. Data analyses
First, baseline characteristics (means [SD], medians [inter quartile range, IQR], numbers [percentages]) were described for all patients who completed the B-IPQ at 3 and 12 months. The differences between responders and non-responders to the 12-months questionnaire were assessed with unpaired t-tests, Mann-Whitney U tests and χ2tests,
de-pending on the distribution of the data. Data analyses were conducted in IBM SPSS v. 24.0 (objectives 1 and 2) and in R v. 3.3.3 (objective 3). For the first objective (description of the IP and their course be-tween 3 and 12 months after the start of rehabilitation) items 1–8 of the B- IPQ at 3 and 12 months after stroke were presented. The means (SD) or medians (interquartile range; IQR) were described, depending on the distribution of the data. Higher values imply more negative IP. To ex-amine the changes of items 1–8 between 3 and 12 months, paired samples t-tests were conducted for each IP. To assess the effect sizes of
the comparisons, Cohen's d values were calculated. Cohen's d values of 0.2, 0.5 and 0.8 were considered small, medium and large respectively [43].
In addition,the numbers (%) of patients whose scores had decreased with one or more points were computed, as well as the numbers (%) of patients whose scores had increased or remained equal (≥0). Additionally, the perceived cause (item 9) that was indicated by the patients as the most important, at 3 months, were coded and categor-ized by two authors (IG and WP) independently. Inconsistencies be-tween the authors were discussed and solved.
For the second objective (identification of clusters of patients with comparable IP trajectories and determine their characteristics), k-means clustering was applied as recommended by Clatworthy et al. [44] In case a patient had a missing IP value, the items was imputed by applying a single imputation with random forests. After that, k-means clustering with a predefined number of 4 clusters was performed using the IP scores at 3 and 12 months. Then, the characteristics of the pa-tients in each cluster were identified and compared using analysis of variance (ANOVA) for normally distributed continuous variables, Kruskal-Wallis test for non-normally distributed continuous variables, and χ2tests for dichotomous variables. Age, sex, stroke hemisphere,
HRQOL physical domain, HRQOL communication domain, depressive symptoms (yes/no), anxiety (yes/no), comorbidities (< 2 vs ≥2) and treatment at 3 months (yes/no) were included in these analyses. In case of significant (p < .05) differences across groups, post hoc analyses were conducted (Games-Howell for normally distributed and Kruskal-Wallis pairwise comparisons for non-normally distributed variables), to which a Bonferroni-corrected significance level of 0.05/3 = 0.017 was applied.
For the third objective (identification of the associations between IP clusters and 12-month mental health), a multivariable logistic regres-sion analysis was conducted with the HADS depressive symptoms (yes/ no) at 12 months as the dependent variable and variables that are likely of influence on depressive symptoms, i.e. IP cluster, age, sex, social support (living with others vs alone) [45], HADS depressive symptoms (yes/no), and HADS anxiety (yes/no) at 3 months as the independent variables. A comparable analysis was conducted for HADS anxiety at 12 months as the dependent variable. ‘Cluster’ was incorporated in the model as two dummy variables, with the largest cluster being the re-ference category.
3. Results
3.1. Study sample and baseline characteristics
Until March 2016 368 patients were invited to the SCORE study of whom 244 (63.2%) signed informed consent. The main reasons not to participate included ‘problems with language’, ‘impaired vision’, or ‘high burden’. One-hundred-and-eighty-four participants (75.4%) pro-vided data on one or more IP 3 months after the start of rehabilitation, and were included in the current study (Fig. 1).
Their mean age was 61.1 (SD 12.7), 107 (58.2%) were male, 144 (78.3%) had an ischaemic stroke, and the average Barthel Index upon admission was 14.2 (SD 5.4). The median number of days between stroke and inclusion into the study was 24.0 (IQR 17.3; 36.0). At 3 months, depressive symptoms were reported by 49 (28.0%) patients and 35 (19.9%) reported symptoms of anxiety (Table 1) Ten patients (5.5%) still stayed at the rehabilitation facility as an inpatient and 92 participants (50%) had treatment as an outpatient. Twelve months after stroke, 151 patients provided data on IP. Among the patients who did not (n = 33), the proportion of higher educated patients and of patients who had finished rehabilitation treatment was significantly higher. 3.2. Description of illness perceptions and the course over time
SD 2.3) and treatment control (mean 3.2, SD 2.5) had the lowest scores (positive) and consequences (mean 6.1, SD 2.8) and timeline (mean 6.0, SD 2.7) had the highest scores (negative). Nine months later, 81 pa-tients had a higher (more negative) score on timeline, meaning that at 12 months patients thought that their disease had a longer duration than they initially thought at 3 months (mean change 0.84, 95%CI 0.36; 1.32). Seventy-nine (56.4%) patients had a higher score on treatment control (mean change 1.42, 95%CI 0.87; 1.97), meaning that at
12 months they perceived the treatment was less effective than they perceived at 3 months. On average, the other IP remained comparable over time, with approximately 33% of patients having a higher and 67% having an equal of lower score at 12 as compared to 3 months. All Cohen's d values were below 0.2, except for timeline (0.28), and treatment control (0.43), which approached a medium effect.
The question on the most important perceived cause of stroke (Table 3) was answered by 110 patients. The causes that were men-tioned most frequently were ‘stress/ worries/ fatigue’ (n = 26, 23.6%), ‘cholesterol, blood pressure, overweight, diabetes’ (n = 25, 22.7%), and ‘lifestyle’ (n = 19, 17.3%).
3.3. Clusters of patients with comparable IP trajectories and their characteristics
Three clusters were discerned of patients with comparable IP tra-jectories, based on their IP scores at 3 and 12 months (Fig. 2a-c ➔ below main text). In cluster 1 (n = 28; 18.8%), the smallest group, the scores at 3 and 12 months were relatively low (positive). In cluster 2 (n = 57; 38.8%) the scores were close to the average of the whole sample. In cluster 3 (n = 64; 43.0%), the scores were relatively high (negative). In all clusters, patients developed a less favourable view on the treatment effect (treatment control) over time. In clusters 2 and 3, patients de-veloped a less favourable view of the duration of the disease but ex-perienced more personal control over time. InTable 4, the differences in patient characteristics across clusters can be found. Significant differ-ences across clusters were found, with the participants in cluster 3 having the most unfavourable scores, for physical functioning (p < .001), communication (p = .003), depressive symptoms (p < .001) and anxiety (p < .001).
3.4. Associations between IP clusters and mental health at 12 months The multivariable logistic regression models revealed that depres-sive symptoms (Exp(B) 6.68, 95%CI 2.13; 20.91) at 3 months was the Questionnaire with ≥ 1 illness
perceptions completed at 12 months: n=151 Included: n=244 (63.2%)
Questionnaire with ≥ 1 illness perceptions completed at 3
months: n=184
Death n=1 Health problems n=2
Withdrawal n=2 Not returned questionnaire n=55
Death n=2 Health problems n=5
Withdrawal n=23 Not returned questionnaire n=3 Invited: n=386
Not willing to participate due to e.g. cognitive problems, impaired vision,
“burden too high”. n=142
Fig. 1. Flow diagram of stroke patients recruited for the SCORE-study until
March 15, 2016.
Table 1
Baseline characteristics of stroke patients included in the analysis of illness perceptions as part of the SCORE-study.
Patients in analyses at
3 months. n = 184 Patients in analyses at 12 months.n = 151 Patients not in analyses at12 months. n = 33 p Sociodemographic characteristics
Age (mean; SD) 61.09 (12.69) 60.58 (12.58) 63.45 (13.09) 0.24
Sex (male; n, %) 107 (58.2) 91 (60.3) 16 (48.5) 0.25
Education (n, %)
Low 79 (45.4) 58 (40.8) 21 (65.5) 0.01
Middle 43 (24.7) 36 (25.4) 7 (21.9) Low vs other: 0.01
High 52 (29.9) 48 (33.8) 4 (12.5)
Living situation (together; n, %) 126 (68.5) 103 (71.5) 23 (74.2) 0.83
Clinical characteristics
Stroke type (ischemic; n, %) 144 (78.3) 119 (78.8) 25 (75.8) 0.82
Stroke localisation (n, %)
Left 84 (45.9) 71 (47.3) 13 (39.4) 0.63
Right 48 (42.6) 64 (42.7) 14 (42.4) Left vs other: 0.45
Other 21 (11.5) 15 (10.0) 6 (18.2)
Barthel Index (mean, SD)
n = 139 14.2 (5.4) 14.2 (5.3) 14.1 (5.6) 0.97
Comorbidities (median, IQR) 1.0 (1.0; 2.0) 1.0 (1.0; 2.0) 1.5 (0.0; 2.8) 0.80
Days between stroke and inclusion, median
(IQR) 24.0 (17.3; 36.0) 24.0 (17.0; 36.0) 24.0 (20.0; 35.5) 0.54
Health status and treatment 3 months post stroke
SAQOL-39NLg physical scale (median, IQR) 4.31 (3.63; 4.81) 4.31 (3.63; 4.79) 4.34 (3.59; 4.81) 0.74 SAQOL-39NLg communication scale (median,
IQR) 4.80 (4.14; 5.00) 4.71 (4.14; 5.00) 4.84 (4.36; 5.00) 0.23
HADS Depressive symptoms (yes; n, %) 49 (28.0) 39 (26.7) 10 (34.5) 0.50
HADS Anxiety (yes; n, %) 35 (19.9) 29 (19.9) 6 (20.0) > 0.99
Treatment at 3 months (in- or outpatient; n, %) 102 (56.0) 91 (61.1) 11 (33.3) 0.01
SD: standard deviation. IQR: inter quartile range. SAQOL-39NLg: Stroke and Aphasia Quality Of Life Scale, HADS: Hospital Anxiety and Depression Scale. NA: not applicable. *Higher = better.
I.F. Groeneveld et al. Journal of Psychosomatic Research 116 (2019) 54–61
strongest predictor for depressive symptoms at 12 months, whereas age, sex, and living situation were not associated with the outcome (Table 5). Moreover, patients in clusters 1 and 2 had a lower risk of depressive symptoms than patients in cluster 3 (Exp(B) = 0.26, 95%CI 0.05; 1.46 and Exp(B) = 0.25, 95%CI 0.07; 0.98, respectively), with the variable ‘cluster 2 vs cluster 3’ being statistically significant. Anxiety at
12 months was significantly associated with depressive symptoms (Exp (B) 4.61, 95%CI 1.33; 15.94) and anxiety (Exp(B) 7.79, 95%CI 2.27; 27.98) at 3 months, but not with any of the other variables. The Na-gelkerke R2of the models were 0.45 and 0.43 respectively.
4. Discussion
We investigated the IP of stroke patients after rehabilitation and its relation to mental health. Three months after the start of inpatient re-habilitation, on average patients believed that the rehabilitation treat-ment would contribute to a large extent to their recovery (treattreat-ment control) and that they understood their disease relatively well (co-herence). However they believed the disease would have a long duration (timeline) and that the disease had a large influence (consequences) on their lives. Nine months thereafter, their perception of disease duration had slightly but significantly deteriorated, and the IP on treatment control had worsened as well. Three clusters were discerned of patients with comparable IP trajectories between 3 and 12 months: Clusters of relatively favourable (1), average (2) and unfavourable IP (3). As ex-pected, the unfavourable cluster was associated with worse physical and mental health at 3 months. Last, a relation between IP and
12-Table 2
Illness perceptions and health outcomes of stroke patients included in the analysis of illness perceptions as part of the SCORE-study, changes between 3 and 12 months after the start of rehabilitation, and the number (%) of patients with equal or improving versus worsening scores.
Illness perceptions 3 months. n = 184.
Mean (SD) 12 months. n = 151.Mean (SD) Change 12–3 months.Mean (95%CI) Cohen's d p Increase (worse), n(%) Equal or decrease(better), n (%)
Consequences 6.1 (2.8) 6.0 (2.7) −0.26 (−0.63; 0.11) −0.11 0.17 50 (33.3) 100 (66.7) Timeline 6.0 (2.7) 7.8 (2.9) 0.84 (0.36; 1.32) 0.28 0.001 81 (55.1) 66 (44.9) Personal control 4.6 (2.7) 4.2 (2.7) −0.46 (−0.97; 0.05) −0.15 0.08 44 (29.7) 104 (70.3) Treatment control 3.2 (2.5) 4.5 (3.2) 1.42 (0.87; 1.97) 0.43 < 0.001 79 (56.4) 61 (43.6) Identity 5.6 (2.6) 5.5 (2.5) −0.11 (−0.47; 0.26) −0.05 0.56 49 (32.5) 102 (67.5) Concern 5.2 (2.9) 5.1 (2.8) −0.29 (−0.79; 0.14) −0.11 0.18 50 (33.3) 100 (66.7) Coherence 3.0 (2.3) 3.0 (2.6) 0.09 (−0.36; 0.54) 0.03 0.71 53 (35.1) 98 (64.9) Emotional response 4.8 (2.9) 4.6 (2.8) −0.18 (−0.60; 0.30) −0.06 0.47 47 (31.8) 101 (68.2)
Physical functioning and
communication 3 months. n = 184.Mean (SD) 12 months. n = 151.Mean (SD) Change 12–3 months.Mean (95%CI) p Decrease (worse), n(%) Equal or increase(better), n (%) SAQOL-39NLg Physical, mean (SD) 4.09 (0.84) 4.14 (0.79) 0.08 (−0.01; 0.16) 0.08 0.09 63 (43.4) 82 (56.6) SAQOL-39NLg Communication,
mean (SD) 4.45 (0.76) 4.41 (0.73) −0.005 (−0.09; 0.08) −0.01 0.91 51 (35.2) 94 (64.8)
Depressive symptoms and anxiety 3 months. n = 184. 12 months. n = 151. Change 12–3 months.
Mean (95%CI) p Equal or increase(worse), n (%) Decrease (better), n(%) HADS Depressive symptoms, mean
(SD) 5.42 (4.10) 5.00 (4.20) −0.30 (−0.84; 0.27) −0.30 0.26 79 (56.4) 61 (43.6)
HADS Anxiety, mean (SD) 5.30 (4.14) 5.01 (3.82) −0.23 (−0.66; 0.19) −0.23 0.28 89 (63.1) 25 (36.9)
SD: standard deviation. SAQOL-39NLg: Stroke and Aphasia Quality Of Life Scale. HADS: Hospital Anxiety and Depression Scale.
Table 3
Most important cause of stroke as perceived by stroke patients included in the analysis of illness perceptions as part of the SCORE-study, 3 months after the start of rehabilitation.
Most important perceived cause of stroke N %
Stress, worries, or fatigue 26 23.6
Cholesterol, blood pressure, diabetes mellitus type 2, or overweight 25 22.7
Lifestyle 19 17.3
Underlying somatic disorder or blood vessel malformation 13 11.8 Healthcare use (malpractice) or medication (side effects) 9 8.2
Genetics 8 7.3
Coincidence or bad luck 8 7.3
Age 2 1.8
Total 110 100
Table 4
The associations between sociodemographic, clinical, 3-month health-related, and treatment-related factors, and clusters of patients with comparable illness per-ceptions trajectories after stroke.
Cluster 1 (favourable)
n = 28 Cluster 2 (average)n = 57 Cluster 3 (unfavourable)n = 64 p omnibus p posthoc
Age, mean (SD) 62.1 (15.0) 61.9 (11.5) 58.8 (12.4) 0.33
Male sex, n (%) 18 (64.3) 37 (64.9) 35 (54.7) 0.46
Living alone, n (%) 7 (25.0) 10 (18.5) 23 (37.7) 0.10
Right hemisphere stroke, n (%) 12 (44.4) 24 (42.1) 28 (43.8) 0.97
≥2 Comorbidities, n (%) 12 (50.0) 16 (38.1) 29 (56.9) 0.36
SAQOL-39NLg Physical scale, median
(IQR)* 4.85 (4.44; 5.00) 4.38 (3.94; 4.75) 3.78 (3.02; 4.44) < 0.001 Cluster 1 vs 2: 0.03 Cluster 1 vs 3:< 0.001 Cluster 2 vs 3: 0.004 SAQOL-39NLg Communication scale,
median (IQR)* 4.86 (4.71; 5.00) 4.86 (4.14; 5.00) 4.43 (3.86; 5.00) 0.003 Cluster 1 vs 2: 0.59 Cluster 1 vs 3: 0.005Cluster 2 vs 3: 0.07 HADS Depressive symptoms (yes = 1) 2 (7.4) 4 (7.4) 32 (50.8) < 0.001 Post hoc tests NA, unbalanced groups HADS Anxiety (yes = 1) 0 (0.0) 4 (7.4) 24 (38.1) < 0.001 Post hoc tests NA, unbalanced groups Treatment (yes = 1) 8 (28.6) 37 (66.1) 46 (73.0) < 0.001 Cluster 1 vs 2: 0.003 Cluster 1 vs 3:
< 0.001 Cluster 2 vs 3: > 0.99
month depressive symptoms was demonstrated. In fact, except for de-pressive symptoms and anxiety at 3 months, IP cluster (2 vs 3) was the only variable significantly associated with depressive symptoms at 12 months.
The results of our study underscored the findings of Grünich et al. on coherence being among the most favourable IP after stroke [19], indicating that patients understood their illness relatively well. Our finding of a declining belief in the effectiveness of treatment (treatment control) in the first year post stroke was also in line with a previous study in stroke [7], as well as in an observational study among dialysis and predialysis patients with chronic kidney failure [46]. Stroke pa-tients may insufficiently realize the persistence of impairments after treatment until their return to normal life, as described in qualitative research [47]. This may also explain their perception of disease dura-tion becoming more unfavourable over time.
As anticipated, patients with more impairments and limitations in physical functioning and communication at 3 months were more likely to be in the unfavourable IP cluster, as they experienced more symp-toms and consequences. Our finding of a relation between IP cluster and depressive symptoms at 12 months was in line with the study of Twiddy et al., who used the General Health Questionnaire to measure distress [7]. Likewise, in breast cancer patients a more negative IP cluster led to more distress after 6 months. [48] Additionally, longitudinal observa-tional studies in patients with various illnesses, such as oesophageal cancer and osteoarthritis, revealed that patients of whom IP worsened over time had higher likelihood of becoming more anxious and de-pressed [25,26].,
As to the perception of the cause of their stroke, a quarter of all patients primarily attributed their disease to stress, worries, or fatigue. The attribution of the disease to chronic stress was previously described in various studies among patients with cardiovascular disease [49]. In a qualitative study among nine transient ischemic attack (TIA) and stroke survivors, 6 patients acknowledged at least one ‘external’ factor such as stress or fate as the cause of their disease, and all 3 patients who mentioned stress considered this as uncontrollable [50]. As described in the literature, the perceived locus of causality (internal or external) and controllability of the disease can influence the patients' coping strategy [51] and may guide future behaviours, including lifestyle and ad-herence to rehabilitation. Informing patients on the controllability of the risk and consequences of stroke may contribute to the prevention of recurrence and adherence to rehabilitation.
From the analyses on the course of IP, both of individual items and within the clusters, it appeared that the items on treatment control and personal control acted different than the other items. This phenomenon was described previously in research among patients with heart disease [52], cancer [53] and multi-morbidity [54]. The study of Timmermans et al. among heart patients, factor analyses on the B-IPQ revealed two factors, i.e. ‘consequences‘ (identity, concern, consequences, emotional response; α=0.80) and ‘control’ (treatment control, personal control,
coherence; α=0.52), with item 2 (timeline) belonging to none of the factors. They found that the total and ‘consequences' scale on the B-IPQ correlated with medical and psychological factors and sex, whereas the ‘control’ scale did not [52]. Therefore they recommend to calculate a total score only for the consequences scale, or analyse all eight items in their own right, which we did in our study.
This is the first comprehensive longitudinal study which describes IP, the changes therein, and the relation with mental health in a sample of stroke patients that is much larger than in previous studies. In con-trast to most other studies using the B-IPQ, the ‘cause’ item was ana-lysed as well. The cluster analyses had several advantages: By making clusters, patients with comparable IP can more easily be characterised in terms of sociodemographic characteristics and health. Moreover, when entering clusters instead of individual items as covariates in a regression model, the risk of type 1 error is reduced as multiple testing is avoided. Several limitations can be mentioned. The first limitation concerns selection bias, which may have distorted the results. Of the stroke patients starting with inpatient rehabilitation, only half partici-pated in the study ánd completed the 3 month questionnaire. Patients with the ‘worst‘ health condition, in terms of cognitive functioning, language, and vision, did not participate. Moreover, the higher edu-cated and those who had finished treatment were more likely to not complete the follow-up questionnaire. Second, the timing of the first assessment was suboptimal; IP were only assessed after the rehabilita-tion trajectory. Thus, we did not provide insight into the ‘added value’ of rehabilitation for influencing IP. Third, a potential limitation of our study is the use of the B-IPQ. Although the concurrent, predictive and discriminative validity were shown to be good, the content validity has been questioned. In each subscale of the B-IPQ only one item was used instead of 4 to 6 as in the original IPQ-R, making it difficult to capture the entire construct. Moreover, a ‘think aloud study’ among 11 patients with various health problems revealed repeated misinterpretations in some of the items of the B-IPQ [55]. However, in that study an adapted version was used in which the instructions to the respondent were unclear, in contrast to our study. In fact, we believe that for stroke patients, the B-IPQ is more feasible and acceptable than the more ex-tensive IPQ-R. As several patients stated in a pilot phase of the SCORE study: extensive questionnaires would be too difficult or tiring to complete.
This study reveals valuable information for health professionals. It became clear that IP of stroke patients are variable over time and that they are related with depressive symptoms. Considering this, IP may be susceptible to CBT and/or self-management interventions. In a sys-tematic review on studies using the B-IPQ, Broadbent et al. described that well designed interventions often succeeded in changing or more IP [34]. For example, in a sample of patients with irritable bowel syn-drome, Chilcot and Moss Morris showed that a CBT self-management program enhanced personal control, facilitated more coherent under-standing of the illness, and reduced perceptions of severe and
Table 5
Results of the logistic regression analyses with depressive symptoms and anxiety at 12 months as dependent variables and patient characteristics, IP cluster, and 3-month depressive symptoms and anxiety as independent variables, in stroke patients participating in the SCORE-study.
HADS depressive symptoms, 12 months HADS anxiety, 12 months
Independent variable Exp(B) 95%CI p Exp(B) 95%CI p
Constant 0.05 0.07 0.08 0.13
Cluster 1 vs 3 0.26 0.05; 1.46 0.13 0.34 0.03; 3.28 0.32
Cluster 2 vs 3 0.25 0.07; 0.98 0.046 1.05 0.27; 4.11 0.35
Age, years 1.02 0.97; 1.06 0.48 1.00 0.96; 1.04 0.95
Sex (1 = male) 1.82 0.62; 5.34 0.28 1.33 0.44; 4.04 0.62
Living situation (1 = together) 1.27 0.41; 3.91 0.68 1.13 0.34; 3.82 0.84
Depression at 3 months (HADS) 6.68 2.13; 20.91 0.001 4.61 1.33; 15.94 0.02
Anxiety at 3 months (HADS) 2.48 0.74; 8.39 0.14 7.97 2.27; 27.98 0.001
Nagelkerke R2 0.45 0.43
HADS: Hospital Anxiety and Depression Scale.
I.F. Groeneveld et al. Journal of Psychosomatic Research 116 (2019) 54–61
distressing consequences of the illness. Moreover they concluded that changes in IP predicted and partially mediated a reduction of symptom severity and an improvement in social adjustment over time [56]. For the rehabilitation setting specifically, these findings are relevant as well. Janssen et al. showed that changes in IP during cardiac re-habilitation were associated with enhanced quality of life [57]. French et al. demonstrated that more favourable IP predicted attendance at cardiac rehabilitation among acute myocardial infarction patients [58]. The effects of CBT and/or self-management interventions in this spe-cific patient group, and the mediating effects of IP on depressive symptoms and other outcomes, should be unravelled in future research. Ultimately, adequately addressing IP in patients with stroke may enrich rehabilitation and prevent the occurrence of depressive symptoms.
5. Conclusion
Illness perceptions partly change over time in the first year after stroke. Patients with worse physical and mental health 3 months after stroke have a less favourable IP trajectory over time. Patients with an overall favourable IP trajectory have a lower risk of depressive symp-toms at 12 months. Health professionals in rehabilitation have a role in the assessment of IP after stroke as it may be an additional target of treatment, most importantly in patients with worse physical and mental health.
Acknowledgements
We are grateful to Inke van Braak and Betsy Nieuwhof for the col-lection of data, to Gerard Volker for his advice in data cleaning, and to Henk Arwert and Radha Rambaran for their role in the SCORE- ad-visory committee.
Funding
This work was supported by the Stichting Kwaliteitsgelden Medisch Specialisten (Medical Specialist Quality Fund; the Netherlands, project nr 328534072014).
Conflict of interest
The authors have no competing interests to report.
Appendix A. Supplementary data
Supplementary data to this article can be found online athttps:// doi.org/10.1016/j.jpsychores.2018.10.019.
References
[1] S. Mendis, Stroke disability and rehabilitation of stroke: World Health Organization perspective, Int. J. Stroke 8 (1) (2013 Jan) 3–4.
[2] S. Geyh, A. Cieza, J. Schouten, H. Dickson, P. Frommelt, Z. Omar, N. Kostanjsek, H. Ring, G. Stucki, ICF Core Sets for stroke, J. Rehabil. Med. (44) (2004 Jul) 135–141 Suppl.
[3] M.L. Hackett, K. Pickles, Part I: frequency of depression after stroke: an updated systematic review and meta-analysis of observational studies, Int. J. Stroke 9 (2014) 1017–1025.
[4] C.A. Campbell Burton, J. Murray, J. Holmes, et al., Frequency of anxiety after stroke: a systematic review and meta-analysis of observational studies, Int. J. Stroke 8 (7) (2013 Oct) 545–559.
[5] S. van Almenkerk, M. Smalbrugge, M.F. Depla, J.A. Eefsting, C.M. Hertogh, What predicts a poor outcome in older stroke survivors? A systematic review of the lit-erature, Disabil. Rehabil. 35 (21) (2013 Oct) 1774–1782.
[6] A. Pollock, G. Baer, P. Campbell, P.L. Choo, A. Forster, J. Morris, V.M. Pomeroy, P. Langhorne, Physical rehabilitation approaches for the recovery of function and mobility following stroke, Cochrane Database Syst. Rev. 4 (2014 Apr 22) CD001920.
[7] M. Twiddy, A. House, F. Jones, The association between discrepancy in illness re-presentations on distress in stroke patients and carers, J. Psychosom. Res. 72 (3) (2012) 220–225.
[8] M. Sjölander, M. Eriksson, E.L. Glader, The association between patients' beliefs
about medicines and adherence to drug treatment after stroke: a cross-sectional questionnaire survey, BMJ Open 3 (9) (2013 Sep 24) e003551.
[9] L.A. Phillips, M.A. Diefenbach, J. Abrams, C.R. Horowitz, Stroke and TIA survivors' cognitive beliefs and affective responses regarding treatment and future stroke risk differentially predict medication adherence and categorised stroke risk, Psychol. Health 30 (2) (2015) 218–232.
[10] C. Harris, Factors influencing return to work after aneurysmal subarachnoid he-morrhage, J Neurosci Nurs. 46 (4) (2014 Aug) 207–217.
[11] H. Leventhal, D. Meyer, D. Nerenz, The common-sense representations of illness danger, in: S. Rachman (Ed.), Contributions to medical psychology, Pergamon Press, New York, 1980, pp. 7–30.
[12] R.R. Lau, K.A. Hartman, Common sense representations of common illnesses, Health Psychol. 2 (1983) 167–185.
[13] R. Moss-Morris, J. Weinman, K. Petrie, R. Horne, L. Cameron, D. Buick, The Revised Illness Perception Questionnaire (IPQ-R), Psychol. Health 17 (1) (2002) 1–16. [14] M.S. Hagger, S. Orbell, A meta-analytic review of the common-sense model of
ill-ness representations, Psychol. Health 18 (2) (2003) 141–184.
[15] E. Broadbent, C.J. Ellis, J. Thomas, G. Gamble, K.J. Petrie, Further development of an illness perception intervention for myocardial infarction patients: a randomized controlled trial, J. Psychosom. Res. 67 (1) (2009 Jul) 17–23.
[16] K.M. Keogh, S.M. Smith, P. White, S. McGilloway, A. Kelly, J. Gibney, T. O'Dowd, Psychological family intervention for poorly controlled type 2 diabetes, Am. J. Manag. Care 17 (2) (2011 Feb) 105–113.
[17] J. Weinman, K.J. Petrie, R. Moss-Morris, R. Horne, The Illness Perception Questionnaire: A new method for assessing the cognitive representation of illness, Psychol. Health 11 (1996) 431–440.
[18] E. Broadbent, K.J. Petrie, J. Main, J. Weinman, The brief illness perception ques-tionnaire, J. Psychosom. Res. 60 (6) (2006 Jun) 631–637.
[19] K. Grünich, V. Garcia-Hoyos, C. Stinear, S. Ackerley, J. Tiemensma, E. Broadbent, Kinematic measures of brain drawings are associated with illness perceptions in people with stroke, Int. Psychogeriatr. 28 (10) (2016 Oct) 1637–1642. [20] S.L. Grace, S. Krepostman, D. Brooks, H. Arthur, P. Scholey, N. Suskin, S. Jaglal,
B.L. Abramson, D.E. Stewart, Illness perceptions among cardiac patients: relation to depressive symptomatology and sex, J. Psychosom. Res. 59 (3) (2005 Sep) 153–160. [21] P. Bower, E. Harkness, W. Macdonald, P. Coventry, C. Bundy, R. Moss-Morris,
Illness representations in patients with multimorbid long-term conditions: qualita-tive study, Psychol. Health 27 (10) (2012) 1211–1226.
[22] P.C. Grayson, N. Amudala, C. McAlear, R.L. Leduc, D. Shereff, R. Richesson, P. Merkel, Illness perceptions and fatigue in systemic vasculitis, Arthritis Care & Research 65 (11) (2013) 1835–1843.
[23] H. De Jong, J. Hillcoat, S. Perkins, et al., Illness perceptions in bulimia nervosa, J Health Psychol 17 (3) (2012) 399–408.
[24] T. Bonsaksen, A. Lerdal, M.S. Fagermoen, Trajectories of illness perceptions in persons with chronic illness: An explorative longitudinal study, J. Health Psychol. 20 (7) (2015 Jul) 942–953 (details).
[25] J. Bijsterbosch, M. Scharloo, A.W. Visser, I. Watt, I. Meulenbelt, T.W. Huizinga, A.A. Kaptein, M. Kloppenburg, Illness perceptions in patients with osteoarthritis: change over time and association with disability, Arthritis Rheum. 61 (8) (2009 Aug 15) 1054–1061.
[26] M. Dempster, N.K. McCorry, E. Brennan, M. Donnelly, L.J. Murray, B.T. Johnston, Do changes in illness perceptions predict changes in psychological distress among oesophageal cancer survivors? J. Health Psychol. 16 (3) (2011 Apr) 500–509. [27] K.J. Petrie, L.D. Cameron, C.J. Ellis, D. Buick, J. Weinman, Changing illness
per-ceptions after myocardial infarction: an early intervention randomized controlled trial, Psychosom. Med. 64 (4) (2002 Jul-Aug) 580–586.
[28] K.M. Keogh, S.M. Smith, P. White, S. McGilloway, A. Kelly, J. Gibney, T. O'Dowd, Psychological family intervention for poorly controlled type 2 diabetes, Am. J. Manag. Care 17 (2) (2011 Feb) 105–113.
[29] E. Jonsbu, E.W. Martinsen, G. Morken, T. Moum, T. Dammen, Change and impact of illness perceptions among patients with non-cardiac chest pain or benign palpita-tions following three sessions of CBT, Behav. Cogn. Psychother. 41 (4) (2013 Jul) 398–407.
[30] S.B. Wang, Y.Y. Wang, Q.E. Zhang, S.L. Wu, C.H. Ng, G.S. Ungvari, L. Chen, C.X. Wang, F.J. Jia, Y.T. Xiang, Cognitive behavioral therapy for post-stroke de-pression: A meta-analysis, J. Affect. Disord. 235 (2018 Aug 1) 589–596. [31] I.F. Groeneveld, J.J. Meesters, H.J. Arwert, A.D. Rambaran Mishre, T.P.M. Vliet
Vlieland, P.H. Goossens, Research design of an analysis of structure, processes and outcomes: Practice variation in stroke rehabilitation, Dutch Journal of
Rehabilitation Medicine 3 (2015) 134–137.
[32] I.F. Groeneveld, J.J. Meesters, H.J. Arwert, P.H. Goossens, T.P.M. Vliet Vlieland, Practice variation in the structure of stroke rehabilitation in four rehabilitation centres in the Netherlands, J. Rehabil. Med. 48 (3) (2016 Mar) 287–292. [33] World Medical Association Declaration of Helsinki, Ethical Principles for Medical
Research Involving Human Subjects, JAMA 310 (20) (2013) 2191–2194. [34] E. Broadbent, C. Wilkes, H. Koschwanez, J. Weinman, S. Norton, K.J. Petrie, A
systematic review and meta-analysis of the Brief Illness Perception Questionnaire, Psychol. Health 30 (11) (2015) 1361–1385.
[35] A.A. Kaptein, I.M. van Korlaar, M.I.P.Q.-K. Scharloo, Translation of B-IPQ in Dutch. Afd. Medische Psychologie, LUMC, 2004.
[36] A.A. Kaptein, IPQ-K,http://www.uib.no/ipq/pdf/B-IPQ-Dutch.pdf. [37] K. Hilari, D.L. Lamping, S.C. Smith, S. Northcott, A. Lamb, J. Marshall,
Psychometric properties of the Stroke and Aphasia Quality of Life Scale (SAQOL-39) in a generic stroke population, Clin. Rehabil. 23 (6) (2009 Jun) 544–557. [38] L. Van Ewijk, L. Versteegde, E. Raven-Takken, K. Hilari, Measuring quality of life in
[39] A.S. Zigmond, R.P. Snaith, The hospital anxiety and depression scale, Acta Psychiatr. Scand. 67 (6) (1983 Jun) 361–370.
[40] I. Bjelland, A.A. Dahl, T.T. Haug, D. Neckelmann, The validity of the Hospital Anxiety and Depression Scale. An updated literature review, J Psychosom Res 52 (2) (2002 Feb) 69–77.
[41] Permanent onderzoek naar de leefsituatie (POLS),http://www.scp.nl/Onderzoek/ Bronnen/Beknopte_onderzoeksbeschrijvingen/ Permanent_onderzoek_naar_de_ leefsituatie_POLSAccessed Sept 5, 2017.
[42] F.I. Mahoney, D.W. Barthel, Functional evaluation: the Barthel Index, Maryland State Med J 14 (1965) 61–65.
[43] J. Cohen, Statistical power analysis for the behavioral sciencies, Routledge, 1977. [44] J. Clatworthy, D. Buick, M. Hankins, J. Weinman, R. Horne, The use and reporting of cluster analysis in health psychology: a review, Br. J. Health Psychol. 10 (2005 Sep) 329–358 Pt 3.
[45] D. Jyotirekha, G.K. Rajanikant, The sequelae of cerebral stroke, Neurosci. Biobehav. Rev. 90 (2018) 104–114.
[46] D.L. Jansen, M.J. Heijmans, M. Rijken, P. Spreeuwenberg, D.C. Grootendorst, F.W. Dekker, E.W. Boeschoten, A.A. Kaptein, P.P. Groenewegen, Illness perceptions and treatment perceptions of patients with chronic kidney disease: different phases, different perceptions? Br. J. Health Psychol. 18 (2) (2013 May) 244–262. [47] R. Wiles, A. Ashburn, S. Payne, C. Murphy, Discharge from physiotherapy following
stroke: the management of disappointment, Soc Sci Med 59 (6) (2004 Sep) 1263–1273.
[48] N.K. McCorry, M. Dempster, J. Quinn, A. Hogg, J. Newell, M. Moore, S. Kelly, S.J. Kirk, Illness perception clusters at diagnosis predict psychological distress among women with breast cancer at 6 months post diagnosis, Psychooncology 22 (3) (2013 Mar) 692–698.
[49] D. French, E. Maissi, T. Marteau, Causal attributions for heart disease a systematic review, Psychol. Health 16 (2001) 77–98.
[50] S. Runions, A. Arnaert, R. Sourial, Causal attributions and health behavior choices among stroke and transient ischemic attack survivors, J Neurosci Nurs. 38 (4 Suppl) (2006 Sep) 288–295.
[51] S.C. Roesch, B. Weiner, A meta-analytic review of coping with illness: Do causal attributions matter? J Psychosom Res 50 (2001) 205–219.
[52] I. Timmermans, H. Versteeg, M. Meine, S.S. Pedersen, J. Denollet, Illness percep-tions in patients with heart failure and an implantable cardioverter defibrillator: Dimensional structure, validity, and correlates of the brief illness perception questionnaire in Dutch, French and German patients, J. Psychosom. Res. 97 (2017 Jun) 1–8.
[53] T. Karataş, Ş. Özen, S. Kutlutürkan, Factor structure and psychometric properties of the brief illness perception questionnaire in turkish cancer patients, Asia Pac. J. Oncol. Nurs. 4 (1) (2017 Jan-Mar) 77–83.
[54] B. Schüz, J.K. Wolff, L.M. Warner, J.P. Ziegelmann, S. Wurm, Multiple illness perceptions in older adults: effects on physical functioning and medication ad-herence, Psychol. Health 29 (4) (2014) 442–457.
[55] L. van Oort, C. Schröder, D.P. French, What do people think about when they an-swer the Brief Illness Perception Questionnaire? A 'think-aloud' study, Br. J. Health Psychol. 16 (Pt 2) (2011 May) 231–245.
[56] J. Chilcot, R. Moss-Morris, Changes in illness-related cognitions rather than distress mediate improvements in irritable bowel syndrome (IBS) symptoms and disability following a brief cognitive behavioural therapy intervention, Behav. Res. Ther. 51 (10) (2013 Oct) 690–695.
[57] V. Janssen, V. De Gucht, H. van Exel, S. Maes, Changes in illness perceptions and quality of life during participation in cardiac rehabilitation, Int J Behav Med. 20 (4) (2013 Dec) 582–589.
[58] D.P. French, A. Cooper, J. Weinman, Illness perceptions predict attendance at cardiac rehabilitation following acute myocardial infarction: a systematic review with meta-analysis, J. Psychosom. Res. 61 (6) (2006 Dec) 757–767.
I.F. Groeneveld et al. Journal of Psychosomatic Research 116 (2019) 54–61
